Purpose: A recent trial called into question the efficacy of breast self-examination. We studied the characteristics and outcome of women in whom physical examination (PE) was their sole method of breast cancer detection.
Patients and methods: From 1970 to 1998, 1752 women with stage I/II breast cancer underwent breast-conserving surgery and radiation. Two hundred sixty patients (15%) had abnormal PE finding as their sole method of cancer detection at the time of diagnosis, 723 (41%) had only mammographic findings, and 762 (43%) had both.
Results: Detection by PE was associated with younger age, larger tumor size, positive axillary nodes, and use of chemotherapy. For women < 40 years of age, PE was the sole method of detection in 40% of cases. The patients for whom PE was the sole method of detection had equivalent 10-year locoregional control and overall survival (OS) compared with patients whose cancer was detected by mammography. Detection by PE was not an independent predictor for outcome on multivariate analysis. The use of tamoxifen (P = 0.0089) was the sole predictor for improved locoregional control. Tumor stage (P = 0.0001), nodal status (P = 0.039), age (P = 0.0112) and lymphovascular invasion (P = 0.0399) were negative predictors of OS.
Conclusion: Although associated with younger age, larger tumors, and more frequent node positivity, in this study detection by PE did not confer worse outcome. This may be because of the increased use of chemotherapy in these patients. Physical examination remains an important method of detection of breast cancer, particularly for younger women for whom mammography is less sensitive and not performed as frequently.